Policy Information Report

America’s Perfect Storm Three Forces Changing Our Nation’s Future This report was written by:

Irwin Kirsch, Henry Braun, and Kentaro Yamamoto of ETS, and Andrew Sum of Northeastern Table of Contents University.

The views expressed in this report are those of the authors and do not necessarily reflect the views of the Preface...... 2 officers and trustees of Educational Testing Service. Acknowledgments...... 2 Additional copies of this report can be ordered for $15 (prepaid) from: Executive Summary...... 3

Policy Information Center. Introduction...... 6 Mail Stop 19-R. Educational Testing Service. A Look Backward...... 8 Rosedale Road. Princeton, NJ 08541-0001. The Three Forces...... 11 (609) 734-5949. [email protected] The Divergent Skill Distributions...... 12

View and download the . The Economy...... 16 report and additional materials . at ets.org/stormreport The Demographic Trends...... 19

A Look Toward the Future...... 21 Copyright © 2007 by . Educational Testing Service. . Implications...... 24 All rights reserved. Educational Appendix...... 27 Testing Service, ETS, and the ETS logo are registered trademarks of . Educational Testing Service (ETS) in the United States of America . and other countries throughout . the world.

January 2007. Policy Evaluation and. Research Center. Policy Information Center. Educational Testing Service

 Preface

Dear God, be good to me; The authors document and describe these three The sea is so wide, forces — divergent skill distributions among U.S. pop- And my boat is so small. ulation groups, a changing economy, and demographic trends of a growing, more diverse population. They project the impact of these interactions upon the na- These lines, from the Breton Fisherman’s Prayer, strike tion 25 years into the future. Kirsch and his colleagues an image of vulnerability for those who may not be warn us that the confluence of these factors can create adequately prepared for a challenging environment. a powerful dynamic that continually feeds the storm A similar image is cast by Kirsch, Braun, Yamamoto, — putting our nation at great risk. They offer hope, and Sum in America’s Perfect Storm: Three Forces however, that if we act now and invest in policies that Changing our Nation’s Future. The authors offer the will help our nation grow together, we can meet our image of our nation as a nautical convoy. Some boats ideals as a nation offering real opportunity for all its are large, well built, and able to ride out the heaviest citizens and continue our leading role in the world. of turbulent seas. Others are smaller, but still quite sturdy, and able to survive. But, many are fragile, mea- gerly equipped, and easily capsized in rough waters. Michael T. Nettles. This convoy is in the midst of a perfect storm that is Senior Vice President. the result of a confluence of three powerful forces. Policy Evaluation and Research Center.

Acknowledgments

The authors wish to acknowledge the following reviewers: the American Federation of Teachers. Amanda McBride Paul Barton, Richard Coley, and Drew Gitomer of ETS; was the editor and Christina Guzikowsky designed the Johan Uvin of the Rhode Island Department of Educa- cover and provided desktop publishing services. Errors tion; and Bella Rosenberg, former Director of Research at of fact or interpretation are those of the authors.

 Executive Summary

Our nation is in the midst of a perfect storm — the in an increasingly competitive work environment. result of the confluence of three powerful forces — that These skills are also needed to function effectively is having a considerable impact on our country. If we in our complex society, with its large bureaucratic maintain our present policies, it is very likely that we institutions and its complex legal, health care, and will continue to grow apart, with greater inequity in retirement systems. wages and wealth, and increasing social and political • More importantly, these skills are not evenly . polarization. If, however, we recognize the power of distributed across groups defined by race/ethnicity, these forces as they interact over the years — and we country of birth, and socioeconomic status. In fact, change course accordingly — then we have an oppor- there are substantial differences in average profi- tunity to reclaim the American dream in which each of ciencies among these groups that influence their us has a fair chance at sharing in any future prosperity. social, educational, and economic opportunities. What are the three forces comprising this perfect • International surveys of student and adult . storm? They are divergent skill distributions, the populations indicate that while our average . changing economy, and demographic trends. performance is no better than mediocre, our degree The first force contributing to our country’s perfect of inequality (the gap between our best and least storm is the wide disparity in literacy and numeracy proficient) is among the highest in OECD countries. skills among our school-age and adult populations. The second force comprises the seismic changes • High school graduation rates peaked at 77 percent in our economy that have resulted in new sources of in 1969, fell back to 70 percent in 1995, and have wealth, novel patterns of international trade, and a stayed in this range into the current decade. The shift in the balance between capital and labor. These graduation rate for disadvantaged minorities is changes have been driven by both technological in- thought to be closer to 50 percent. A recent report novation and globalization, resulting in a profound by the Organisation for Economic Co-operation restructuring of the U.S. workplace. Indeed, the labor and Development (OECD) indicates that the United markets of today are markedly different from those of States ranked 16th out of 21 OECD countries with earlier decades. For example: respect to high school graduation rates.1 • In 1950, manufacturing’s share of total employment • Data from the National Assessment of Educational in the United States was 33.1 percent. By 1989, it was Progress (NAEP) reveal that between 1984 and down to 18.2 percent and, by 2003, it was 10.7 percent. 2004 reading scores among 13- and 17-year-olds • Between 1984 and 2000 the number of employed remained flat, and the achievement gaps were large persons 16 years of age and older grew by 29 . and relatively stable. For mathematics the story is percent, or some 30 million. At the same time, only slightly different. While the mean scores for employment in jobs associated with college-level both the nation’s 13- and 17-year-olds improved education grew by some 20 million, accounting for slightly, they did so across all groups, with the result two-thirds of the job growth. that the average size of the Black-White and His- panic-White achievement gaps remained large and • The country’s employment growth is expected to relatively stable. continue through the rest of this decade and into the next, with college labor market clusters . • National surveys of our adult population indicate (professional, management, technical, and high- that large numbers of our nation’s adults, 16 years level sales) expected to generate about 46 percent . of age and older, do not demonstrate sufficient liter- of all job growth between 2004 and 2014. acy and numeracy skills needed to fully participate

1 A number of alternative methodologies have been developed by educational researchers to estimate high school graduation rates. For a discussion and review of various approaches see Gary Orfield (ed.), Dropouts in America: Confronting the Graduation Rate Crisis, . Cambridge, MA, Harvard Education Press, 2004.  One important consequence of this shift in the com- • In 2004, nearly 57 percent of the 16- to 64-year-old position of jobs in our country has been the increasing Hispanic population in the United States was for- economic returns to schooling and skills. For example: eign-born, up from 46 percent in 1990. More than half of these immigrant Hispanics lacked a high • The expected lifetime earnings of males with a bach- school diploma. elor’s degree in 1979 were 51 percent higher than their peers with only a high school diploma. By 2004, • The lack of a high school diploma by such a large however, this difference had widened to 96 percent. proportion of Hispanic immigrants is of concern given the fact that almost 80 percent of immigrants • The earnings premiums accruing to a particular who have not earned a high school diploma report level of educational attainment (e.g., high school not speaking English well or at all. diploma, bachelor’s degree) are substantially larger for individuals at that level who have higher cogni- These three forces — substantial disparities in the tive skills, indicating that both education and skills distributions of skills, economic restructuring, and contribute to individual opportunities. These oppor- demographic trends — are each powerful in their tunities include not only higher paying jobs but also own right. But as they play out together over time, the the chance for individuals to take advantage of em- result is truly a perfect storm that, unlike the storm ployer-sponsored training to enhance and broaden chronicled by Sebastian Junger2, continues to gain their skills throughout their working lives. strength with no end in sight. The third force involves sweeping demographic • Employing demographic projections combined changes. The U.S. population is projected to grow with current skill distributions, we estimate that by from nearly 300 million in 2005 to more than 360 mil- 2030 the average levels of literacy and numeracy in lion in 2030. Over this period, our population will be- the working-age population will have decreased by come increasingly older and more diverse, with immi- about 5 percent while inequality will have increased gration having a significant impact on the composition by about 7 percent. Put crudely, over the next 25 of the workforce, as well as of the general population. years or so, as better-educated individuals leave the workforce they will be replaced by those who, on • The U.S. labor force is projected to grow more average, have lower levels of education and skill. slowly over the next 20 years than it did between Over this same period, nearly half of the projected 1980 and 2000. None of this growth is predicted to job growth will be concentrated in occupations as- come from native-born workers of prime working sociated with higher education and skill levels. This age (25 to 54). means that tens of millions more of our students • During the 1980s, international migration account- and adults will be less able to qualify for higher- ed for about 21 percent of our nation’s population paying jobs. Instead, they will be competing not growth; with that contribution rising to 31 percent only with each other and millions of newly arrived in the 1990s. Moreover, the U.S. Census Bureau . immigrants but also with equally (or better) skilled expects that between 2000 and 2015, net interna- workers in lower-wage economies around the world. tional migration will account for more than half. As we argue in this report, it is both the growth of of our nation’s population growth. human capital and how it is distributed that is impor- • Fueled both by higher birth rates and by immigra- tant for the United States. Human capital is critical, tion, the Hispanic share of the population is expect- however, not just because of the economic implica- ed to grow from 14 percent in 2005 to slightly more tions. Benjamin Friedman3 links our economic well- than 20 percent by 2030. being to the fabric of our society, arguing that individ- uals and societies are more trusting, more inclusive,

2 Sebastian Junger authored The Perfect Storm: A True Story of Men Against the Sea. Published in 1997, Junger’s book recounts the tale of the October 1991 “perfect storm,” focusing on the loss of the Gloucester sword-fishing boat Andrea Gail off the coast of Nova Scotia. 3 See Benjamin M. Friedman, “Meltdown: A Case Study,” Atlantic Monthly, July/August 2005.

 and more open to change when they see their futures and that of their children as bright and secure. He . concludes that “Economic growth is not merely the enabler of higher consumption; it is in many ways the wellspring from which democracy and civil society flow.” • Given the forces described in this report, a looming question is whether we will continue to grow apart or, as a nation, we will invest in policies that will help us to grow together. We strongly believe the . latter is the better course of action over the long term. • While new policies focusing only on education and skills will not solve all the challenges associated with existing inequalities, if our society’s overall lev- els of learning and skills are not increased and the existing gaps are not narrowed, there is little chance that economic opportunities will improve among key segments of our population.

 Introduction

On a continent bounded by two oceans, our society has The second set of forces is driven by the fact that often employed nautical metaphors to generate evoca- the U.S. economy has undergone significant changes tive images. National political leaders used to speak, in recent years. The labor markets of today are mark- for example, of the “ship of state,” and President John edly different from those of earlier decades as a result F. Kennedy, in arguing the case for economic growth of several developments, including industrial and in the early 1960s, claimed that “a rising tide lifts all corporate restructuring, declines in unionization, boats.”4 Given our country’s growing demographic technological change, and globalization. The last set diversity, however, perhaps it would be more appro- of changes were made possible by the combination of priate now to imagine our nation as a convoy. Some a series of international trade agreements, the rapid of the boats are large, well built, and able to ride out development and diffusion of information and com- the heaviest of seas. Others are somewhat smaller, munications technology (ICT) infrastructure, advances less well-equipped, but still quite sturdy. But many in logistics, and foreign investment in developing are fragile, meagerly equipped, and easily swamped in countries. An important consequence of technologi- rough waters. That convoy — the individuals, families cal change and globalization is increasing economic and communities that make up our nation — is in the returns to higher levels of education and skills.8 For midst of a “perfect storm,” the result of the confluence workers who possess these characteristics, opportuni- of three powerful sets of forces: divergent skill distribu- ties abound. But for those who lack them, the econom- tions, a changing economy, and demographic trends. ic future can be quite dismal. The first set of forces is evident in the findings of The third set of forces is powered by demographic a number of large-scale national and international trends. The U.S. population is projected to grow from surveys conducted over the past decades, which show nearly 300 million in 2005 to more than 360 million that large numbers of our nation’s adults do not have by the year 2030. Certain minority groups, especially sufficient command of the literacy and numeracy5 Hispanics, represent a considerable and growing pro- skills needed to function effectively in an increasingly portion of the population. Immigration has accounted competitive work environment. Moreover, skills are for an increasingly large fraction of U.S. population not evenly distributed across groups defined by race/ growth over the past few decades and is projected to ethnicity, nativity,6 and socioeconomic status. In fact, do so over the remainder of this decade — and at least there are substantial differences in average literacy into the near future. According to the U.S. Census Bu- levels among these groups. In younger cohorts, these reau, these immigration patterns, coupled with differ- differences are mirrored by critical differences in ences in birth rates, will result in substantial increases academic achievement and educational attainment. in the racial/ethnic diversity of the nation’s population Moreover, comparisons made possible by a number over the next 20 to 25 years and beyond. For example, of international assessments, including PISA (Pro- Hispanics, who now constitute about 14 percent of the gramme for International Student Assessment) and population, will constitute more than 20 percent by IALS (International Adult Literacy Survey), indicate 2030; and non-Hispanic African Americans, who now that our younger cohorts are losing ground relative to constitute some 12.5 percent of the population, will their peers in many developed nations.7

4 President Kennedy used this expression on a number of occasions during the 1960s. The earliest cited time was during the presidential campaign of 1960. 5 The construct of numeracy was used in the Adult Literacy and Life Skills (ALLS) survey and was defined as the knowledge and skills . required to effectively manage and respond to the mathematical demands of diverse situations. 6 The term nativity is used to distinguish those who were born in the United States from those born outside of the United States. 7 See Organisation for Economic Co-operation and Development, Learning for Tomorrow’s World: First Results from PISA 2003, Programme for International Student Assessment, Paris, 2004, and Andrew Sum, Irwin Kirsch, and Robert Taggart, The Twin Challenges of Mediocrity and Inequality: Literacy in the U.S. from an International Perspective, Policy Information Report, Center for Global Assessment, Policy Infor- mation Center, ETS, 2002. 8 See Richard J. Murnane and Frank Levy, Teaching the New Basic Skills, New York, The Free Press, 1996. For a review of the economic im- pacts of these basic skills on the weekly and annual earnings of workers in the United States, see Andrew Sum, Literacy in the Labor Force, National Center for Education Statistics, Washington, D.C., 1999, and Andrew Sum, Irwin Kirsch, and Kentaro Yamamoto, Pathways to Labor Market Success: The Literacy Proficiency of U.S. Adults, Policy Information Report, Policy Information Center, ETS, 2004.

 then constitute nearly 14 percent. Over that same pe- On the other hand, if we respect the storm’s power riod, Asian Americans will see their share of the popu- and change course appropriately, we still have a lation increase from 4 percent to almost 7 percent. chance to help more Americans share in any future prosperity. Surely, our nation will be different than the This is not the first report to take note of one or one that sailed into the storm more than a decade ago more of these forces, or to warn of the challenges we — but it will retain many of the characteristics that face as a nation. Some of the reports discussed in have distinguished it from those of other countries. If, the next section also talk about our mediocrity and as a nation, we agree that strengthening the convoy inequality in literacy and numeracy skills, as well as and preserving its distinctiveness are important, then the strong associations among education, literacy, and (perhaps) we will act more in concert with our ideals economic success.9 This report uses data from recent of a nation having real opportunity for all. national and international surveys to report on the skill distributions of our school-age and adult popula- tions. It also looks at shifts taking place in the work- force, and at demographic projections made by the U.S. Census Bureau. Using statistical methodologies, we project what the literacy skills of the next genera- tion will look like if projected changes in our demo- graphics are allowed to interact with existing skill distributions. While critics of the earlier reports sometimes sug- gest that their rhetoric exaggerates America’s challeng- es, our analyses indicate that their rhetoric is not over- blown. We are in the midst of a perfect storm in which these three powerful forces are combining to generate waves that already have had a considerable impact on our nation. Unlike the perfect storm chronicled in the novel written by Sebastian Junger, the forces behind this storm continue to gain strength, and calm seas are nowhere in sight. We can’t hope to ride this one out. If we continue on our present heading and fail to take . effective action, the storm will have a number of pre- dictable and dire implications for future generations, with consequences that extend well beyond the economic realm to the ethos of our society.

9 Mediocrity and inequality refer to the performance of the United States on international surveys relative to other highly industrialized nations. These terms are taken from the title of a report written by Andrew Sum, Irwin Kirsch and Robert Taggart, The Twin Challenges of Mediocrity and Inequality: Literacy in the U.S. from an International Perspective, Policy Information Report, Policy Information Center, ETS, February 2002.

 A Look Backward

Historically, the United States has provided eco- pre-eminent position in the global economy. The nomic opportunity for most of those willing to work United States led most of the industrialized world in hard, invest in themselves, and act responsibly. For the percentage of adults with a high school and college the most part, we still subscribe to that promise, and education and accounted for a disproportionate share both the civil rights and women’s rights movements of world gross domestic product. have extended this promise to groups often neglected Several other post-World War II events are worthy in previous decades. As the nation changed from a of note. The launch of Sputnik in 1957, as well as the predominantly agrarian society to an industrial one apparent growing economic power of the Soviet Union in the late 19th and early 20th centuries, labor-market in the 1950s and early 1960s, alerted the nation to success increasingly depended on attaining at least a the need for strengthening mathematics and science modest level of formal education, along with a willing- education. The “War on Poverty,” initially declared by ness to “do a day’s work for a day’s pay.” Over time, as President Johnson in 1964, committed the federal gov- the structure of jobs and the economy changed, occu- ernment to reducing poverty partly by improving edu- pational and basic literacy skills became more essential cation investments in the poor from pre-school (Head for obtaining a decent job and advancing in the work- Start) through adulthood. The Civil Rights Act of 1964 force. Eventually, a high school diploma emerged as a represented an attempt to use federal legislation to spur key credential for economic success. Through much of greater equality in U.S. society — in part through provid- the 20th century, the proportion of high school gradu- ing more equitable access to educational opportunity.11 ates,10 as well as the proportion of adults who gradu- ated from college, increased with each successive age Over the past two decades, a series of reports have cohort — although those proportions differed substan- been published that focus on the links between educa- tially by race/ethnicity and, in some cases (e.g., college tional performance, literacy skills, and the U.S. econ- education), by gender. omy. In 1983, the National Commission on Excellence in Education (sometimes referred to as the Bell . After World War II, the GI Bill of Rights provided Commission) issued a report on the state of the . the initial impetus for a rapid expansion in college . nation’s educational system. In that report, the . attendance and, eventually, the number of college commission made the following widely cited statement graduates. Although this policy was motivated by about the quality of our school system: domestic policy considerations, it was certainly con- sistent with the American ethos and led to the develop- The educational foundations of our society ment of a stronger human capital base that helped to are presently being eroded by a rising tide of drive productivity and economic growth. Not only did mediocrity that threatens our very future as a the education and skill levels of U.S. workers increase nation and a people. What was unimaginable during this period, so did the number of individuals a generation ago has begun to occur — others who were entering the workforce. Thus, businesses are matching and surpassing our educational had a growing base of both men and women who attainments ... we have allowed this to happen were better-educated and better-skilled than those they to ourselves.12 replaced. This unprecedented fund of human capital, Shortly following the publication of A Nation at together with the fact that the United States (unlike Risk, the Committee for Economic Development is- Europe and Japan) had escaped the worst ravages of sued its own report, Strategy for U.S. Industrial Com- World War II, enabled this country to achieve a . petitiveness, which called for a variety of economic,

10 See Paul Barton, One-Third of a Nation: Rising Dropout Rates and Declining Opportunities, Policy Information Report, Policy Information Center, ETS, February 2005. In this report, Barton notes that the percentage of high school graduates as a percentage of our nation’s 17- year-olds peaked at 77 percent in 1969, dropped to 70 percent in 1995, and has stayed close to that level at least through 2001. 11 Unlike previous initiatives, the Civil Rights Act was an explicit attempt to address the social and economic inequalities that continue to exist in our country. 12 National Commission on Excellence in Education, A Nation At Risk: The Imperative for Educational Reform, U.S. Government Printing . Office, Washington, D.C., 1983.  educational, and technological reforms to boost the Concerns with slow productivity growth, real-wage future competitiveness of the U.S. economy.13 The re- stagnation, rising wage inequality, and international port identified the quality of education, particularly at competitiveness continued throughout the remainder the primary and secondary levels, as a key determinant of the 1980s. In 1990, the National Center on Educa- of the future ability of the U.S. labor force to adjust tion and the Economy published a report on the work- to workplace changes that would be produced by new force-development challenges facing the American technologies designed to improve labor productivity. economy if it wished to maintain high wages for the future workforce.16 The report identified a number of In 1986, the Carnegie Corporation’s Forum on Edu- critical choices the nation must make about the strate- cation and the Economy published a report document- gies it would pursue to bolster its economic competi- ing the changing structure of the economy, its likely tiveness. The Commission on the Skills of the Ameri- impacts on the education and skill requirements of can Workforce, co-chaired by two former Secretaries jobs, and the consequences for the nation’s schools and of Labor, William Brock and Ray Marshall, called for a the teaching profession. In A Nation Prepared: Teachers wide array of actions by national, state, and local gov- for the 21st Century, the commission argued that ernments and private-sector firms to increase invest- If America wants to compete on the same terms ments in the education and training of both youth and as it did in the past — making the most of the adults to allow the country to maintain a high-produc- workers with low skill levels — then it must ac- tivity/high-wage trajectory. cept prevailing world wage levels for low-skilled In 2003, the Hoover Institution’s Koret Task Force and semi-skilled labor. That is we must be on K-12 Education issued its report on an appraisal of prepared for a massive decline in our standard the state of the nation’s educational system, updating of living. The alternative is to revise our view findings that originally appeared in A Nation at Risk.17 of the role of the worker in the economy. In the Their report was titled Our Schools and Our Future: future, high-wage level societies will be those Are We Still at Risk? The limited progress of the nation whose economies are based on the use of a wide since the initial report, with respect to both improv- scale of very skilled workers, backed up by the ing high school graduation rates and strengthening most advanced technologies available.14 the literacy and numeracy skills of our students, was The growing importance of basic academic skills cited by the task force. In their final report, the Koret (reading, math, and writing) in determining an array Task Force described the accelerated consequences of of educational, labor market, and social outcomes for globalization since the publication of A Nation at Risk: America’s youth was identified in a 1988 monograph The shrinking globe has made it easier than published by the Ford Foundation’s Project on Social anyone in 1983 could have imagined for invest- Welfare and the American Future.15 In Toward a More ments and jobs to go anywhere on the planet Perfect Union, the authors identified important associa- that seems likeliest to succeed with them. Here tions between basic academic proficiencies of youth we must look to our schools to produce the and their school performance, graduation rates, college highly educated citizenry on which America’s attendance rates, childbearing behavior, experience future economic vitality depends.18 with the criminal justice system, and annual earnings.

13 Committee for Economic Development, Strategy for U.S. Industrial Competitiveness, New York, 1984. 14 Carnegie Forum on Education and the Economy, A Nation Prepared: Teachers for the 21st Century, Carnegie Corporation of New York, . New York, 1986. 15 Gordon Berlin and Andrew Sum, Toward a More Perfect Union: Basic Skills, Poor Families, and Our Economic Future, Ford Foundation Project on Social Welfare and the American Future, New York, 1988. 16 Commission on the Skills of the American Workforce, America’s Choice: High Skills or Low Wages!, National Center on Education and the Economy, Rochester, New York, 1990. 17 See Koret Task Force on K-12 Education, Our Schools and Our Future: Are We Still at Risk? Hoover Institution, Stanford University, 2003. 18 A condensed version of the Koret Task Force report also appeared in the following article: “Our Schools, Our Future: Are We Still at Risk?” in Education Next, Spring 2003, pp. 9-15.

 In November 2005, the National Center for Pub- recently in 2002 with the passage of the No Child Left lic Policy and Higher Education warned that un- Behind Act. This legislation explicitly addresses the less states made significant efforts to improve the importance of having every child meet state-man- educational attainment of all residents, the personal dated standards of proficiency and sets out a system of income of Americans would decline over the next 15 school and district accountability measures to encour- years.19 Similarly, the National Academy of Sciences age compliance. While educational analysts often issued its own report, warning the nation that it was disagree over the types of reforms needed to achieve “on a losing path” in global economic competition.20 these objectives, there is general agreement that too The report, titled Rising Above the Gathering Storm, many children are leaving school too early, and too contained a series of recommendations to bolster the many of those who persist to earn (say) a high school nation’s long-run economic competitiveness, including diploma emerge with weak skills and insufficient a strengthening of the math and science proficiencies knowledge.21 of the nation’s K-12 students and increased national As the next section of this report will show, these commitments to basic research and the training and warnings and concerns about our educational system recruitment of scientists. and its influence on wages and the economy are well- In response to these and other concerns, federal grounded. Unless we are willing to make substantial involvement in education over the past three decades changes, the next generation of Americans, on average, has evolved through successive reauthorizations of will be less literate and have a harder time sustaining the Elementary and Secondary Education Act, most existing standards of living.

19 The National Center for Public Policy and Higher Education, Income of U.S. Workforce Projected to Decline if Education Doesn’t Improve, Policy Alert, November 2005. 20 Committee on Prospering in the Global Economy of the 21st Century: An Agenda for American Science and Technology, Rising Above the Gathering Storm, National Academy of Sciences, Washington, D.C., 2005. 21 For a review of alternative views on the problems of our schools and the need for reform, see the Spring 2003 issue of Education Next.

10 The Three Forces

In the following sections, we analyze each of the three Finally, we use information provided by the U.S. sets of forces that constitute America’s perfect storm. Bureau of the Census to examine recent and projected First, we examine the divergent skill distributions, us- demographic changes. According to these projections, ing data from recent national and international sur- the U.S. population will continue to grow, but will be- veys. Skill distributions have a prominent role in this come increasingly older and more diverse. Immigration report, not only because they are strongly associated will continue to have a significant effect on both the with key social and economic outcomes of adults, but general population and those who enter the workforce. also because they have been shown to be associated According to David Ellwood, America’s labor force will with these outcomes over and above their effects on grow much more slowly over the next 20 years than it educational attainment. Differences in the distribution did between 1980 and 2000, with almost none of the of both educational attainment and skills by major growth coming from native-born workers of prime subgroups in our population mirror the variation in working age (i.e., 25 to 54 years).22 Between 2000 and earnings, income, and wealth seen in American society. 2005, two-thirds of the nation’s civilian labor-force growth and 86 percent of its employment growth was Next we summarize key aspects of the economic generated by new immigrant arrivals. forces that are changing the nature of work for many Americans. There has been a shift in the composition of jobs toward those employing individuals with high- er levels of education and skills. This shift is predicted to continue through the remainder of this decade and well into the next, with important implications for who is most likely to gain access to these jobs.

22 See David Ellwood’s contribution to the report Grow Faster Together or Grow Slowly Apart: How Will America Work in the 21st Century? . The Aspen Institute, 1998.

11 The Divergent Skill Distributions

One of the implications of a globalized, technologi- increasingly complex society. Performance in Levels cally-driven economy is that the notion of what charac- 3 and higher is considered to be a minimum standard terizes high-skilled workers is changing as well. Murnane for success in the labor market.26 Findings from the and Levy have written about the “new basic skills,” National Adult Literacy Survey (NALS) indicate that in which, they argue, comprise a broader range of both 1992 slightly more than half of the U.S. adult popula- cognitive skills (the traditional three R’s) along with a set tion ages 16 to 65 years of age performed in Level 3 of the so-called “soft skills.”23 In fact, one can argue that or higher on each of the three scales. A decade later, most critical are those foundational skills that enable . results from the Adult Literacy and Life Skills Survey individuals to learn throughout their lifespan and, thus, be (ALLS) showed that the percentages decreased to be- able to adapt to changing work conditions and demands. low 50 percent.27 On that score, we are no- where near where we need to be as a nation. Over the past Figure 1 15 years, a number of literacy Percentage of Adults in Each Prose Literacy surveys of nationally representa- Proficiency Level, by Race/Ethnicity tive samples of adults have been 24 carried out. In these surveys, 60 adult proficiencies were mea- All Adults Asian Black Hispanic White sured along three distinct scales 50 49 (prose literacy, document litera- 44 cy, and quantitative literacy [nu- 41 meracy]).25 For each proficiency 40 scale, adults were categorized 35 32 33 33 33 into one of five levels, with Level 30 30 30 28 1 being the lowest and Level 5

the highest. Adults in Levels 1 Percentage 20 20 20 and 2 are characterized as pos- 17 13 14 sessing, at best, basic- or inter- 12 mediate-level skills, respectively. 10 9 Although few of these adults 3 3 would be considered illiterate 0 1 2 3 4/5 1 2 3 4/5 1 2 3 4/5 1 2 3 4/5 1 2 3 4/5 in the historical meaning of the Low High Low High Low High Low High Low High term, only a small percentage Prose Literacy Level were judged to have the skills needed to fully participate in an Source: Adult Literacy and Life Skills (ALLS) Survey, 2005

23 Murnane and Levy, 1996. 24 These surveys include the National Adult Literacy Survey (NALS, 1992), the International Adult Literacy Survey (IALS, 1995), the Adult Literacy and Life Skills Survey (ALLS, 2005), and the National Assessment of Adult Literacy (NAAL, 2005). 25 NALS and IALS employed a quantitative literacy scale, but ALLS employed a numeracy scale instead. While the quantitative scale dealt primarily with arithmetic skills embedded in texts, the numeracy scale focuses more on mathematical reasoning and, as such, represents a broader construct that is more relevant to technologically advanced societies. Relatively speaking, adults performed least well on the numeracy scale in the ALLS survey. 26 A number of national and state organizations in the United States, including the National Governor’s Association, identified proficiency in Level 3 as a minimum for success in today’s labor market. See, for example, John Comings, Andrew Sum, and Johan Uvin, New Skills for a New Economy: Adult Education’s Role in Sustaining Economic Growth and Expanding Opportunity, Massachusetts Institute for a New Commonwealth, Boston, December 2000. 27 See Irwin Kirsch, Ann Jungeblut, Lynn Jenkins, and Andrew Kolstad, Adult Literacy in America: A First Look at Results of the National Adult Literacy Survey, U.S. Department of Education, September 2003. See also Organisation for Economic Co-operation and Develop- ment and Statistics Canada, Learning a Living: First Results of the Adult Literacy and Life Skills Survey, Paris, 2005.

12 An equally disturbing picture emerges when the profi- Asian adults. Readers interested in more detail on these ciencies of adults from various subgroups of the popula- literacy data are referred to Appendix Tables A1 and A2. tion are compared. Figure 1, using ALLS data, shows Does it really matter where one falls on these lit- the percentages of adults ages 16 to 65 in each prose eracy/numeracy scales? Yes. In fact, data from these proficiency level by race/ethnicity. According to the ALLS same surveys reveal substantial differences by literacy data, Black, Hispanic, and Asian adults are significantly level in labor-force participation rates, job skill-level more likely to perform in the lowest level on the prose requirements, weekly and annual earnings, and access scale when compared with White adults. The percentage to lifelong learning. For example, only 49 percent of of Hispanic adults who demonstrate skills in the lowest adults in Level 1 on the prose scale reported partici- of the five literacy levels is four times that of White adults pating in the workforce (i.e., being either employed or (49 percent compared with 12 percent). Among Black unemployed and looking for work). This percentage adults, this ratio is 2.8 times, while for Asian adults it is increased to 77 percent for adults in Level 3, and to 91 2.3. Not surprisingly, therefore, much smaller percentag- percent for those in Level 5. Similar trends were found es of these three racial/ethnic groups demonstrate skills for the document and quantitative/numeracy scales. in the highest two levels on the prose scale. Here we see 17 percent of White adults perform in Levels 4 and These differences in literacy skills and labor-market 5 compared with only 3 percent of Black and Hispanic participation rates also are accompanied by consider- adults and 9 percent of Asian adults. able gaps in earnings.28 In 1992, average weekly earn- ings were about $355 for adults in Level 1, about $530 Additional insight into the sources of these differ- for those in Level 3, and about $910 for those in Level ences can be gained by considering these subgroups 5 on the prose scale. That is, the weekly earnings of further divided by nativity status. In 2003, according to adults in Level 5 were some 2.6 times those of adults the U.S. Bureau of the Census, foreign-born individuals in Level 1. Furthermore, adults with higher literacy/ represented some 12 percent of the U.S. population, with numeracy skills were more likely to participate in for- about 25 percent coming from Asia and more than 50 mal education and training than those with more lim- percent coming from Latin America. When we compare ited skills. The ALLS report noted that adults in Level the average differences in literacy scores with White 3 were nearly four times more likely to participate adults of the same nativity, there are some interesting in these types of programs compared with adults in patterns. Non-native Black adults perform better on the Level 1, with this ratio increasing to almost six times prose scale than their native-born peers — the average for adults in Levels 4 and 5 on the prose scale. Thus, skills gap among native Black adults compared with . not only are less-skilled adults less able to adapt to a native White adults is three times larger than it is among changing employment climate, they are less likely to non-native Black and White adults (42 points compared have the opportunity to develop the requisite skills.29 with 14 points on the prose scale). In contrast, among Hispanics, the skills gap between non-native Hispanic One might argue that while concerns about today’s and White adults almost doubles, increasing from an gaps in literacy and numeracy are understandable, average of 32 points to an average of 61 points. Yet an- the current focus on educational improvement, the other pattern holds for Asians: Native-born Asian adults institution of national and state standards, along with outscore native-born White adults by more than 30 increased spending on K-12 education, ensure that the points on the prose scale, while non-native White adults gaps are closing. Unfortunately, recent reports on high perform only slightly better (11 points) than non-native school graduation rates30, as well as on student per-

28 Estimates are based on those persons who reported full-time employment (i.e., working at least 35 hours per week). 29 Sum, Kirsch, and Yamamoto, 2004, Organisation for Economic Co-operation and Development and Statistics Canada, 2005. 30 A number of alternative methodologies have been developed by educational researchers to estimate high school graduation rates. For a comprehensive review of alternative estimates of high school dropout rates in the United States, and variations across regions, states, and demographic groups, see Orfield, 2004; Andrew Sum et al., The Hidden Crisis in the High School Dropout Problems of Young Adults in the U.S.: Recent Trends in Overall School Dropout Rates and Gender Differences in Dropout Behavior, report prepared by the Center for Labor Market Studies, Northeastern University, Boston, for the Business Roundtable, Washington, D.C., 2002.

13 formance in reading and mathematics, suggest other- content area, performance is described on measures wise. The 2001 public high school on-time graduation that range from 0 to 800, with an overall mean of 500. rates for Black (50 percent) and Hispanic (53 percent) On the composite math scale, for example, the students were well below those of White (75 percent) mean score of U.S. 15-year-olds was 483, which was and Asian (77 percent) students.31 International com- 17 points below the OECD average score of 500. The parisons offer no comfort: For calendar year 2003, the difference is statistically significant, with the United United States ranked 16th out of 21 OECD countries States ranked 24th-highest among the 29 OECD with respect to high school graduation rates. Our countries for which data were available. In fact, the graduation rate, estimated to be 73 percent, was well U.S. score was significantly lower than those of 20 below the 90 percent or more reported for Germany, countries, statistically the same as three countries, and Greece, Ireland, Japan, Norway, and Switzerland.32 significantly above only five OECD countries. On the Performance data from national surveys are no science assessment, the U.S. mean score was 9 points more promising. Data from the National Assessment below the OECD average. This performance merits a of Educational Progress (NAEP) indicate that be- rank of 20th among 29 OECD countries. On the read- tween 1984 and 2004 average reading scores of both ing assessment, the U.S. results were more nearly in the nation’s 13-year-olds and 17-year-olds have been the middle of the pack. The value added by instruction essentially flat. During this same period, the Black- through age 15 for many students seems to be consid- White and Hispanic-White achievement gaps have erably lower than in many other industrial countries.35 remained large and relatively stable. In math, the story is somewhat brighter. Between 1986 and 2004, average math scores in NAEP increased significantly for both 13-year-olds and 17-year-olds. The growth in math was Table 1 shared among all groups of students, but the average Comparative Performance of U.S. 15-Year-Old size of the Black-White and Hispanic-White achieve- Students on the 2003 PISA Math, Reading, and 33 ment gaps remained large and relatively stable. There Science Assessments is also considerable evidence that these gaps are sub- stantial even for students attending the same schools.34 Math Reading Science Readers interested in more detail on NAEP scores are U.S. Average Score 483 495 491 referred to Appendix Tables A3 and A4. OECD Average 500 494 500 Looking beyond our borders, data from PISA reveal U.S.-OECD Difference -17 +1 -9 that in no area of achievement is the United States among the world’s leaders in terms of average educa- U.S. Rank Among 24th 15th 20th tional output or yield at age 15 — despite the fact that 29 Countries we are among the leaders in per-pupil spending. Select- Source: PISA, 2003. ed results from the mathematics, reading, and science assessments for 2003 are displayed in Table 1. In each

31 Christopher B. Swanson, “Sketching a Portrait of Public High School Graduation: Who Graduates? Who Doesn’t” in Dropouts in America: Confronting the Graduation Rate Crisis, Cambridge, MA, Harvard Education Press, 2004. 32 Organisation for Economic Co-operation and Development, Education at a Glance, OECD Indicators, 2005. 33 Marianne Perie, Rebecca Moran, and Anthony D. Lutkus, NAEP 2004 Trends in Academic Progress: Three Decades of Student Performance in Reading and Mathematics (NCES 2005-464), U.S. Department of Education, Institute of Education Sciences, National Center for Educa- tion Statistics, Washington, D.C., July 2005. 34 Henry Braun et al., “The Black-White Achievement Gap: Do State Policies Matter?” Education Policy Analysis Archives, 14 (8), March 20, 2006. Retrieved 4-20-06 from http://epaa.asu.edu/epaa/v14n8/. 35 Herbert J. Walberg, “Examinations for Educational Productivity,” in Williamson Evers and Herbert J. Walberg (eds.), Testing Student Learning, Evaluating Teacher Effectiveness, Hoover Institution Press, Stanford University, Stanford, 2004; and Herbert J. Walberg, Spending More While Learning Less: U.S. School Productivity in International Perspective, Thomas B. Fordham Foundation, July 1998.

14 This mediocre performance is not limited to our The combination of our relative (mediocre) position school-age populations. Data from the IALS shown with respect to average performance and our leading in Table 2 provide a basis for comparing U.S. adults position with respect to inequality in performance to adults in 19 other participating high-income coun- leads to concern about the growing danger to the well- tries. Only on the prose scale, where the United States being of our nation. This disparity in skills is related to ranked 9th highest, was the U.S. score significantly the disparity we see in the educational attainment of higher than the mean for all 20 high-income countries. our population and to the growing disparities in social Again, on none of the literacy measures is the United and economic outcomes. Ignored, these differences States a world leader. In fact, on each of the three may not only reduce our ability to compete interna- scales, the U.S. average is approximately one-third to tionally on a high-wage strategy, but also will surely one-half of a standard deviation below the two interna- threaten the cohesiveness of the nation. tional frontrunners: Sweden and Norway. On the IALS, as on PISA, the U.S. average is just that — average. Table 2 We are among the world leaders, however, in the amount of variability in our score distributions — the Comparison of Mean Scores on IALS for the gap between our best and worst performers. In IALS, United States and 20 High-Income Countries the United States was ranked either first or second High-Income on this measure. Similarly, in PISA 2000, the United U.S. Countries U.S. Rank States was among the countries with the highest Prose Literacy 273 267 9th standard deviation on the overall reading scale. Only Document Literacy 267 267 13th Belgium, Germany, and New Zealand were reported to Quantitative Literacy 274 270 13th have larger standard deviations. Source: IALS, tabulations by authors, 2002.

15 The Economy

The reorganization of work in the United States under ing not only of manufacturing jobs but also of so-called competitive pressures, and the impetus of technologi- knowledge worker jobs is becoming increasingly com- cal changes (particularly the introduction of informa- mon.37 From call centers and help desks to software tion technologies), has had — and will continue to have development, income tax form preparation, financial — a substantial impact on the structure of jobs and research, engineering and architectural design, and the demand for certain skill sets in our labor market. medical diagnosis, it is both feasible and economically Computerization and globalization have changed attractive to have work performed where wages and manufacturing enormously. To cite but one example, the cost of benefits are much lower than in the United technological developments within the industry, such States. In Thomas Friedman’s memorable phrase, “the as computer-controlled lathes, automated assembly world is flat.”38 That is, barriers to the flow of work have lines, and robotic quality control, have altered auto- been significantly reduced, with both positive and nega- mobile production so that significantly fewer — but tive effects on U.S. workers and consumers, with the higher skilled — workers are needed to produce a fixed former facing downward pressures on wages and the number of vehicles. These shifts have been accelerated latter benefiting from lower prices. by international trade agreements that have increased There also has been a dramatic shift in the com- the share of imports in the industry, further reducing position of the job distribution in America toward domestic employment. In other areas, such as textiles, professional, management and management support, large international wage differentials and dramatic technical, and high-level sales positions.39 Many of the reductions in the cost of shipping make it possible faster-growing private-sector services and financial to send raw materials halfway around the world, use industries employ a greater proportion of individuals them to manufacture garments, and return the com- in these college-labor-market occupations. At the same pleted garments to the United States at a fraction of time, changes in technology and the mix of industries what it would cost to manufacture the same garments in manufacturing have also increased the demand for on our shores. workers with these skills. For example, in 1984, only At the same time, large, structural trade deficits 21 percent of those employed in the nation’s manufac- mirror the dramatic losses in employment in the turing industries held a job in a professional, mana- nation’s manufacturing industries. Between 1999 and gerial, or technical occupation.40 By 2000, nearly 28 2004, payroll employment in the nation’s manufactur- percent of workers in manufacturing were employed ing industries declined from 17.3 million to slightly in these occupations. under 14.4 million, a drop of nearly 17 percent. From 1984 to 2000, the number of employed per- Looked at over a longer period, the situation is even sons age 16 and older (wage and salary workers, as worse. In 1950, manufacturing’s share of total employ- well as self-employed workers) increased by more than ment was 33.1 percent. By 1989, it was down to 18.2 30 million. This represented an increase of nearly 29 percent and, by 2003, it was 10.7 percent.36 percent over the 16-year period. Within each of the More recently, with the advent of substantial management, professional, technical, and high-level amounts of worldwide broadband capacity, outsourc- sales categories, job growth exceeded that for all U.S.

36 Eamonn Fingleton, Unsustainable: How Economic Dogma is Destroying American Prosperity, New York, Nation Books, 2003. 37 For a review of outsourcing developments and their impacts on U.S. workers, see C. Alan Garner, “Offshoring in the Service Sector: Eco- nomic Impact and Policy Issues,” Federal Reserve Bank of Kansas City, Economic Review, 3rd Quarter, pp. 5-37, and Timothy Sturgeon and Frank Levy, Measuring the Offshoring of Service Work and Its Impact on the United States, MIT, Cambridge, MA, March, 2005. 38 Thomas Friedman, The World Is Flat: A Brief History of the 21st Century, New York: Farrar, Straus, and Giroux, 2005. For a more nuanced view of Friedman’s book, see Edward Leamer, A Flat World, a Level Playing Field, a Small World, After All, or None of the Above? Available at uclaforecast.com/reviews/Leamer_FlatWorld_060221.pdf. 39 Under the occupational coding system used by the U.S. Census Bureau in the CPS survey until 2003, occupations were classified on this 1990 Census occupational coding structure. From 2003 onward, the U.S. Census Bureau shifted to the new SOC Occupational Classifica- tion System that utilizes a different system for categorizing jobs into occupational clusters. We have combined SOC categories for the 2000-2005 and 2004-2014 periods to closely match the earlier concepts of management, professional, and technical positions. 40 See U.S. Bureau of Labor Statistics, Employment and Earnings, January 1985, Table 25, p. 183.

16 workers. Employment growth in these four college- These changes in the occupational structure in the labor-market occupation clusters exceeded 20 million; United States have important consequences for the that is, they accounted for two-thirds of all of the job educational and literacy requirements of jobholders. growth in the nation even though they represented The demand for college-educated and more highly lit- only one-third of all of the nation’s workers at the erate workers has increased over the past two decades, start of this period in 1984. The two fastest-growing enhancing the economic returns to schooling and occupational categories were management-related skills. The literacy and numeracy proficiencies of U.S. positions, which increased by nearly 71 percent, and adults, as well as their educational attainment, have a professional employment, which increased by nearly strong influence on their ability to gain access to the 59 percent, both more than doubling the average for more highly skilled professional, management, and all occupations.41 technical occupations — the job categories that also provide more opportunities to apply reading, writing, Over the 2000 to 2005 period, civilian employment and math skills as well as to participate in employer- is estimated by the U.S. Bureau of Labor Statistics sponsored training. to have risen by only some 5 million, or 3.7 percent. Employment in management-related, professional, and The proportion of employed adults who held profes- technical occupations combined grew by 2.9 million, sional, management-related, technical, or high-level or 6.4 percent — more than double the growth rate for sales occupations at the time of the 1992 NALS was all other occupations in the labor market.42 The profes- strongly associated with their completed years of post- sional, management-related, and technical occupations secondary schooling and their prose and quantitative generated just under 60 percent of the net growth in proficiency scores. Among both associate degree and employment between 2000 and 2005. If high-level sales bachelor’s degree recipients, the likelihood of an em- workers are included in the total, then their share of net ployed person holding a college-labor-market job was job growth comes closer to 65 percent.43 strongly and consistently correlated with their prose and quantitative proficiencies (see Table 3). National employment projections by the U.S. . Bureau of Labor Statistics foresee a strengthening of Interestingly, college graduates with weak literacy job growth over the 2004 to 2014 period, with the total and numeracy proficiencies were much more likely number of employed persons increasing by some 19 than their more highly skilled peers to be underem- million, or 13 percent. If these projections hold true, ployed — holding jobs that did not require a college then professional, management, technical, and high- degree and paid considerably less than their more level sales positions will generate about 46 percent of highly skilled counterparts. For example, less than half all job growth in the nation over the 2004-2014 de- of employed four-year college graduates whose prose cade.44 Of the remaining mix of jobs in our economy, proficiency placed them in Level 1, and only a slight only service occupations, many of which are relatively majority of those in Level 2, were working in college- low-skilled, are projected to grow at an above-average labor-market occupations. pace (19 percent). Employment of clerical/administra- The ALLS survey found that U.S. workers with tive support and production workers will increase at a strong skill profiles were much more likely to have rate well below the average for all occupations. Readers jobs that required them to engage in reading and writ- interested in more data on changes in the employment ing tasks at work.45 In particular, the survey revealed structure are referred to Appendix Tables A5, A6, and A7.

41 The occupational employment estimates for calendar year 2000 are based on the use of population weights from the 1990 Census. These employment estimates were later revised by the Bureau of Labor Statistics (BLS) to reflect new population weights from the 2000 Census, but they did not revise the occupational employment estimates under this coding system. 42 Between 2000 and 2005, the rate of growth for all other occupations was only 2.6 percent. 43 In the SOC classification system, we consider sales supervisors and sales representatives as high-level sales workers, but we exclude real estate brokers. 44 For a more detailed review of the employment outlook for the nation, see U.S. Bureau of Labor Statistics, “Employment Outlook: 2004-14,” Monthly Labor Review, November 2005. 45 Organisation for Economic Co-operation and Development and Statistics Canada, Learning a Living: First Results of the Adult Literacy and Life Skills Survey, Paris, 2005. 17 Table 3 Percentage of All Employed Workers and Two-Year and Four-Year College Graduates with Jobs in Professional, Technical, or Management Occupations, by Prose Proficiency Level, 1992

Schooling Level 1 Level 2 Level 3 Level 4 Level 5 Overall Two-Year Degree 28% 29% 37% 43% 40% 38% Four-Year Degree or Higher 46 56 64 75 83 71 All Employed Workers 5 14 26 50 72 27

Source: Andrew Sum, Literacy in the Labor Force: Results from the National Adult Literacy Survey, 1999. very strong links between the literacy proficiencies of information that comes our way. As the likelihood of workers and their access to knowledge-expert, mana- long-term employment declines, and as greater num- gerial, and high-skill information-processing jobs.46 bers of individuals will be required to assume more re- For example, only 13 percent of U.S. adults were at sponsibility for managing various aspects of their lives Levels 4 or 5 on the numeracy scale, but they repre- (including career planning, health care, and retire- sented 36 percent of the workers in knowledge-expert ment), higher levels of skills will be required for full jobs and one-fifth of those in high-skill information- participation in our society. The growing divergence in processing occupations. Nearly three-fourths of the labor-market outcomes and prospects based on educa- workers in low-skill service occupations had numeracy tion and skills, and the implications these trends have skills at or below Level 2. for our society, deserve serious attention from policy- makers at all levels. The average weekly pay gap between workers in knowledge-expert and management positions and Figure 2 those in low-skill service occupations is quite size- able, with the former earning a positive premium of Estimated Weekly Earnings Premiums of 68 percent, and the latter a negative premium of 50 Workers in Selected Occupations in the United percent relative to the estimated weekly earnings of States Relative to Manufacturing-Related manufacturing-related occupations (Figure 2). These Occupations (in Percentages) differential premiums translate into a salary ratio of 80 more than 3-to-1. 68

In fact, the relative wage gap in the United States 60 57 between knowledge experts or managers and low- skill service workers was the second highest of the six 40 countries studied in the 2003 ALLS assessment. Only 21 Canada’s was larger. 20 We can speculate as to the cause and effect of the increasing return on education and skills that we have Percent 0 seen in recent years; that is, whether it has resulted from a greater demand for credentials to gain access -20 to existing jobs or from actual changes in education -20 and skill requirements of jobs. Regardless, higher -40 levels of skill and education are important not only for gaining access to better jobs, but also for negotiating -50 -60 our bureaucratic society and its complex legal, health Knowledge Managers Information Information Service care, and retirement systems, and for accessing and Experts High Skills Low Skills Low Skills comprehending the seemingly limitless amount of Source: OECD and Statistics Canada, Learning a Living, 2005.

46 Organisation for Economic Co-operation and Development and Statistics Canada, Learning a Living, “Chapter Six: Skills and the Nature of the Workplace,” Paris, 2005. 18 The Demographic Trends

The U.S. population is diverse and becoming more gration. Mexico has been by far the largest source of so. In recent years, foreign immigration to the United new immigrants, with the 2.17 million arrivals from States has reached levels not seen in almost a century. that country constituting one-third of the total. Only This immigrant population has contributed at an in- one of the other 12 countries, Canada, was a predomi- creasingly high rate to the nation’s population growth nantly English-speaking country. Although a number of and, especially, to growth in the nation’s civilian labor the highly educated immigrants from several of these force.47 During the 1980s, net international migration countries did study English before coming to the United accounted for only 21 percent of the nation’s population States, relatively few immigrants arrive in the United growth, but its contribution to population growth rose to States with a strong base of English-speaking, -reading, 31 percent in the 1990s.48 The U.S. Census Bureau projects and -writing skills. that, over the 2000-2015 period, net international migra- New immigrants to the United States are quite tion will continue to increase in both absolute terms and heterogeneous with respect to both their educational as a percentage of the nation’s overall population growth. attainment and their English-speaking skills. As shown In fact, it will likely account for more than half of the in Table 4, among new immigrants age 18 and older nation’s population growth during that period. who were living in the United States in 2004, approxi- From April 2000 to July 2005, the U.S. popula- mately one-third lacked a high school diploma while tion grew by about 15 million, with net international some 28 percent held a bachelor’s or higher academic migration of 6.3 million constituting about 42 percent degree. Not surprisingly, the educational attainment of of the growth. The impact foreign immigration has these new immigrants varied considerably by region had on the labor force and on employment growth of origin. While 62 percent of the immigrants from over the same five years has been even greater because Mexico and Central America lacked a high school di- (1) a large proportion of foreign immigrants are of ploma, only 9 percent of the immigrants from Europe working age and (2) foreign immigrants, particularly lacked a high school diploma. A majority (56 percent) men, participate at a relatively high rate in the nation’s labor force. It is estimated that between 2000 and Table 4 2005 new immigrants contributed two-thirds of the The Distribution of New Immigrants (Age 18 and nation’s labor force growth and nearly 86 percent of older) into the United States, by Educational the country’s net employment growth, including more Attainment, 2000-2004 than 100 percent of the growth in jobs held by males. Clearly, as these numbers illustrate, foreign immigra- Number tion constitutes an increasingly important factor in the Educational Attainment (thousands) Percent U.S. economy (see Appendix Table A8). 1 to 12 years, no diploma 1,671 34 12 years, high school diploma 1,147 23 As one might expect, new immigrants to the United or GED States over the 2000 to 2004 period came from al- 13 to 15 years 777 15 most every locale around the globe. A large propor- Bachelor’s degree 828 17 tion, however, came from a relatively small number of countries (see Appendix Table A9). The 13 countries Master’s degree or higher 556 11 sending 100,000 or more immigrants during this Total 4,979 100 period accounted for about two-thirds of all immi- Source: 2004 American Community Surveys, public use files. Tabulations by the authors.

47 For a recent review of the effects of new foreign immigration on population, labor force, and employment growth in the United States since 2000, see Andrew Sum, Paul Harrington, and Ishwar Khatiwada, New Foreign Immigrants and the United States Labor Market: The Unprecedented Effects of New Foreign Immigration on the Growth of the Nation’s Labor Force and Its Employed Population, 2000-2004, pa- per presented to the U.S. Congress, House Subcommittee on Immigration, Border Security and Claims, May 4, 2005. 48 Net international migration represents the difference between the flow of new immigrants into the United States and the flow of U.S. residents (including former immigrants) to other countries. The 1980s estimates were derived from The Statistical Abstract of the United States, 1993-94. Reference Press, Austin, Texas, 1994. The 1990s estimates of net international migration were based on U.S. Census Bureau estimates for the April 1990 – July 1999 period; but they were made before the 2000 Census findings, which revealed a larger total and immigrant population. 19 of the new immigrants from Asia held a bachelor’s or native-born counterparts on recent literacy and . higher degree, and 46 percent of those from Europe numeracy assessments. For example, on the 1995 did so. However, only 6 percent of the immigrants IALS, the average proficiency scores of U.S. immi- from Mexico and Central America held a four-year . grants were typically more than one standard devia- college degree or a more advanced degree. tion below those of their native-born counterparts on each of the three literacy scales.49 The median score The self-reported English-speaking proficiencies of of immigrants on the composite literacy scale ranked these new immigrants also varied quite widely, with only at the 17th percentile on the U.S. test distribution 32 percent reporting that they either spoke English or on that same scale. spoke it well, while about half claimed that they either did not speak English at all (23 percent) or not well As noted at the outset, the Hispanic population is (26 percent). The English-speaking skills of these new projected to grow substantially over the next 25 years. immigrants were strongly associated with their for- By 2030, its share of the U.S. population will increase mal educational attainment, as shown in Table 5. For from 14 percent to some 20 percent. This growth is example, nearly 8 of 10 new immigrants lacking a high being fueled both by higher birth rates and by elevated school diploma reported that they either could not immigration rates. According to the findings of the speak English at all or could not speak it well. Among American Community Surveys, in 2004, nearly 57 per- those with a master’s degree or higher, however, only cent of the 16- to 64-year-old Hispanic population in 15 percent indicated an inability to speak English or the United States was foreign-born, up from 46 percent speak it well. in 1990. Slightly more than half of these immigrant Hispanics lacked a high school diploma. By contrast, Not surprisingly, the formal educational attain- only slightly more than one-fourth of their native-born ment of immigrants and their English-speaking skills counterparts lacked a high school diploma. The high were very strong predictors of their scores on national proportion of immigrants among the Hispanic adult literacy assessments, including the IALS surveys. population, coupled with the limited schooling and Overall, with their more limited years of schooling and English-speaking proficiencies of many of these immi- weak English-speaking proficiencies, the immigrant grants, accounts for the relatively weak performance of population in the United States scored well below their Hispanics on national literacy assessments.

Table 5 Distribution of New Immigrants into the United States, Age 16 and Older, by Self-Reported English-Speaking Proficiency and by Educational Attainment, 2000-2004

Speaks Does Not Does Not Only Speaks English Very Speaks Speak English Speak Educational Attainment English Well English Well Well English 1 to 12 years, no diploma 6% 6% 9% 32% 47% 12 years, high school diploma or GED 9 16 19 36 21 13 to 15 years 16 30 24 21 9 Bachelor’s degree 14 36 28 17 5 Master’s degree or higher 11 48 27 12 3 Total 10 22 19 26 23

Source: 2004 American Community Surveys, public use files. Tabulations by the authors.

49 Sum, Kirsch, and Yamamoto, 2004.

20 A Look Toward the Future

The three sets of forces we have described — highly . The story is similar for the document and quantita- divergent skill distributions; labor-market restructuring, tive scales. On each of the literacy scales, the average including the forces of globalization; and demographic score is predicted to decrease about 5 percent, and the trends — are each powerful in their own right. But . variability of the distribution is expected to increase by as they interact over time, their consequences can be . about 6 to 8 percent.51 One consequence of this change truly momentous. Let’s try to track the trajectory of . in the distribution of skills is that there will be tens of this perfect storm. millions more adults (in Levels 1 and 2) who lack the education and skills they will need to thrive in the new Of course, predicting the future is fraught with economy. They will be less competitive with equally or uncertainty. However, trends in globalization and basic better-skilled workers in lower-wage economies, and demographic projections provide a reasonably solid will not be well prepared to compete for higher-pay- foundation for prognostication. Although we are less ing jobs in our economy. Obtaining access to higher certain about future trends in overall skill levels, as well earning opportunities as opposed to obtaining jobs will as the distribution of skills for different segments of become the more formidable challenge. the population, let’s assume that the skill distributions within racial/ethnic and age groups remain much the Figure 3 same as they are now. While this may seem pessimistic, it is consistent with recent observations and it will, at The Distribution of Prose Literacy Proficiency from the very least, provide us with a plausible scenario. the 1992 NALS Compared with the Distribution Projected for 203052 As we have already noted, two minority groups that will see increases in their proportions of the popula- 1992 NALS tion between now and 2030 are Hispanics and non- 2030 Projection Hispanic Blacks. Both groups lag considerably behind non-Hispanic Whites and Asians in their average literacy and numeracy skills. We can combine the cur- rent skill distributions by racial/ethnic groups and age groups with demographic projections to predict what the distribution of skills for the nation as a whole will look like in 2030. Figure 3 shows the distribution of prose literacy proficiency as it was in 1992 and as it is projected to be in 2030. Changes are shown for each level in prose, < Level 1 Level 1 Level 2 Level 3 Level 4 Level 5 document, and quantitative literacy in Figure 4 (also, 175 225 275 325 375 425 see Appendix Table A10). It is clear from these graphs Prose Proficiency Distribution that the distribution of skills is expected to shift to the Source: Data for 1992 from National Adult Literacy Survey, 1992; data for 2030 projected by left, corresponding to a substantial reduction in the the authors. Data are for adults ages 16 to 65. percentage of adults in Levels 3 and 4, with a con- comitant — and dramatic — increase in the percent- age of adults at or below Level 1, which jumps from These literacy projections arise from the confluence 17 percent to 27 percent. Equally troubling is the fact of two of the forces: the existing skill distributions and that the distribution of prose proficiency becomes less the projected shift in demographics. What about the peaked at Level 3 and more dispersed overall.50 third force? How will changing proficiency distribu-

50 The levels on each of the literacy scales are defined as follows: Level 1 = 0 to 225; Level 2 = 226 to 275; Level 3 = 276 to 325; Level 4 = 326 to 375; Level 5 = 376 and higher. 51 An increase in variability corresponds to an increase in inequality in the sense that, relatively speaking, there are more individuals at the low and high ends of the scale. 52 While the graph focuses on prose literacy, similar results were found for the document and quantitative literacy scales. These data for all three scales are shown in Figure 4 on the following page and Table A10 in this report’s Appendix.

21 Figure 4 Change in the Distribution of Prose, Document, and Quantitative Literacy, 1992 and 2030 (Projected)

40 1992 2030

35 34 33

30 28 28 28 28 27 27 27 26 26 26 26 25

20 19 19 19 18 17 17 16

Percentage 14 14

10

5 4 4 4 3 2 0 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 Low High Low High Low High Prose Literacy Levels Document Literacy Levels Quantitative Literacy Levels

Source: Data for 1992 from National Adult Literacy Survey, 1992; data for 2030 projected by authors. Data are for adults, ages 16 to 65. tions interact with the continuous economic restruc- important to recognize, however, that individuals with turing that is placing a premium on higher literacy the same level of educational attainment can have very and numeracy skills? What are the likely consequences different literacy skills. Since earnings are influenced by for the nation’s future economic and social structure? both educational attainment and skills, increasing the former without increasing the latter will likely not result A recent report issued by the National Center for in the desired outcomes. Public Policy and Higher Education cites U.S. Census data indicating that, in 2000, White adults ages 25 to Further, recent data from the National Longitudi- 64 were twice as likely as African American adults to nal Survey of Youth (NLSY 1979) and the National have a bachelor’s degree and almost three times as Education Longitudinal Study of 1988 reveal that lit- likely as Hispanic adults.53 According to the report, eracy/numeracy skills are a critical factor in determin- these differences in educational attainment, coupled ing who in the long run earns a bachelor’s or higher with substantial increases in the proportions of African degree. Analyses of these data by one of the authors of American and Hispanic adults comprising the working- this report reveals that almost 76 percent of 36- to 43- age population, could result in a less-educated workforce year-olds in the 10th (top) decile of the basic academic over the next 20 years. One notable consequence would skills distribution from the NLSY 1979 cohort earned be a decline in personal income per capita across the a bachelor’s or higher degree, compared with only 15 United States in the absence of strengthened investments percent of those in the 6th decile and only 2 percent in the schooling of these groups.54 of those in the lowest decile.55 This strong relation- ship between academic skills and ultimate educational That report calls on the states to do a better job of attainment holds regardless of gender or racial/eth- improving the educational levels of all residents. It is nic background (see Appendix Table A11).56 Formal

53 The National Center for Public Policy and Higher Education, Income of U.S. Workforce Projected to Decline if Education Doesn’t Improve, Policy Alert, November 2005. 54 That is, crudely speaking, as better-educated White non-Hispanics leave the workforce over the next two decades, increasingly their places will be filled by workers from minority groups with lower levels of education. 55 Analyses of the National Longitudinal Survey of Youth 1979 cohort. Data reported here and shown in Appendix Table A11 are from the 2002 survey round. 56 It is interesting to note that, at the lower deciles, the likelihood of earning a bachelor’s or higher degree favors women over men, and . minority groups over Whites. That is, except at the highest decile, Blacks are more likely to obtain a college degree than White non-. Hispanics within the same decile (i.e., having similar levels of cognitive skills).

22 schooling and literacy/numeracy skills clearly are be- among segments of our population, especially among coming the great sorter in American society. As former key racial/ethnic groups. President Clinton noted in a 1995 address to a national Trends in family formation in the United States also convention of newspaper editors: “... the technology play a role in rising inequality in education and family revolution and the global economy. These are dividing income. Findings from the 2004 American Community opportunity at home and abroad. The middle class is Surveys, shown in Figure 5, indicate that at least 44 splitting apart, and the fault line is education.”57 percent of all births to women under age 30 were out of Indeed, differences in educational levels and initial wedlock. Among Black mothers in this age group, this skills translate into gaps in annual earnings that, over proportion reaches 77 percent. The vast majority of these the past two decades, have been widening. The lifetime out-of-wedlock births have taken place among women earnings differences between adults with a bachelor’s with either no high school diploma or with no postsec- degree and those with only a high school diploma ondary education (see Appendix Tables A16 and A17). widened considerably between 1979 and 2004 (see This trend in single-parent families, if it continues, will Appendix Table A12). As young adults age, the relative lead to ever-greater family income inequality, in view of size of the gap in the mean annual earnings between the low earnings potential of many of these mothers. high school and college graduates grows considerably. Moreover, children raised in low-income, single- Analysis of longitudinal data from the NLSY 1979 parent families often suffer from a number of critical survey reveals that the mean annual earnings of 23- to cognitive, health, and nutritional deficits that are likely 30-year-old adults with 16 or more years of school- to limit their future academic achievement and edu- ing in 1988 were 58 percent higher than those of high cational attainment. To return to our metaphor, the school graduates. Nine years later, the gap in mean result will be an increase in the number and relative earnings between these two groups had widened to proportion of small, poorly equipped boats in our fleet 81 percent — and would increase to 101 percent by that will find it difficult to thrive — let alone survive 2001. By 2001, those adults (now 36 to 43 years of age) — in the turbulent seas ahead.59 with 16 or more years of schooling had mean earnings twice as high as those of their peers with only a high school diploma. Perhaps more notable is the fact that Figure 5 the premiums for education are substantially greater Percentage of Out-of-Wedlock Births to Women for individuals with higher scores on a measure of cog- Under Age 30, by Racial/Ethnic Group, 2003-2004 nitive skills than for individuals with lower scores.58 All 44 These and other data reveal that both educational Asian 16 attainment and skills are strongly and positively as- Black 77 sociated with annual earnings and access to the more Hispanic 46 highly skilled professional and management positions Mixed Race in the U.S. labor market (for additional data, see Ap- 60 pendix Table A14). Thus, while reducing inequalities in White 34 education and skills will not solve all the challenges . 0 20 40 60 80 100 associated with existing economic inequalities, without Percentage narrowing the overall skills and education gaps there is Source: 2004 American Community Surveys, public use files. little chance of improving the economic opportunities Tabulations by Center for Labor Market Studies. Note: Out-of-wedlock births are modestly underreported on the ACS surveys.

57 The Boston Globe, Saturday, April 5, 1995, p. 1. 58 For example, for individuals who scored near the bottom of the Armed Forces Qualification Test (AFQT) distribution, the premium for a college degree over a high school diploma was 14 percent in 1998, rose to 18 percent in 1997 and to 29 percent in 2001. By comparison, for individuals who scored near the top of the AFQT distribution, the premium was 69 percent in 1988 and 122 percent in 2001. The data can be found in Appendix Table 13. 59 Data reported here are from the 2004 American Community Surveys. They were tabulated by one of the authors of this report and are shown in Appendix Tables A15, A16, and A17.

23 Implications

Continuous industrial restructuring, technologi- important factor in achieving continuous economic cal changes in the workplace, as well as accelerated growth and in reducing the current inequalities in outsourcing of services that were previously insulated earnings and wealth overall and among recognizable from the forces of international competition mean that subgroups. Unfortunately, current skill gaps coupled more and more types of jobs, including professional with demographic trends portend diminishing human and management-support jobs, will be up for grabs in capital among the future prime-working-age popula- a global marketplace. To make matters worse, less- tion of the United States.61 educated U.S.-born workers also face increased com- Accordingly, we must set a national goal of equip- petition from newly arrived immigrants with limited ping most (if not nearly all) adults with the ability to schooling and skills. To prosper in this new labor-mar- perform work that is highly valued in the marketplace ket environment, individuals must possess the skills, and the capability to periodically renew themselves as knowledge, flexibility, and credentials that will allow the nature of that work evolves. In fact, with secure, them to compete successfully. long-term employment becoming rarer and society If we continue on our current course, however, it is becoming more complex, there is a growing need for likely that our nation will gradually lose ground in . all individuals to become better educated and more relation to other countries, becoming more divided skilled. Individuals must have a broad set of literacy both socially and economically in the process. In a and numeracy skills to successfully manage the vari- recent Wall Street Journal editorial, former Treasury ous aspects of their lives: from planning their careers, Secretary Robert Rubin argued for changing our eco- to nurturing and guiding their children, to navigat- nomic strategies — in part to achieve a prosperity that ing the health-care system, to taking more personal is more widely shared, stating: “Broad participation in responsibility for their financial future. economic well-being and growth is critical, both as a fun- Why is the current distribution of educational at- damental value and to realize our economic potential.”60 tainment and academic achievement so problematic If we are unable to substantially close the existing — quite apart from the large and persistent racial/eth- skill gaps among racial/ethnic groups and substantially nic gaps? Early dropouts, as well as many holders of boost the literacy levels of the population as a whole, high school diplomas or GEDs, typically lack strong demographic forces will result in a U.S. population in foundational skills and the key noncognitive skills 2030 with tens of millions of adults unable to meet the that would enable them to benefit from learning and requirements of the new economy. Moreover, a sub- training experiences later on in life. In fact, the lack of stantial proportion of those adults will be members of both cognitive and the so-called “soft skills” (e.g., per- disadvantaged minority groups who will likely con- sistence, teamwork skills, and communication skills) sider themselves outside the economic mainstream. makes it more difficult for them to obtain and retain Clearly, this is not the preferable direction for the . employment, especially for those jobs offering the . nation to be moving. potential for advancement. The employment they do find is unlikely to offer them opportunities to improve Of course, a nation’s economic competitiveness is their skills either on the job or in formal training . not solely determined by the efficiency and effective- programs funded by the employer. ness of its education system, or the literacy skills of its workforce. Many other factors, including the health In contrast, individuals at the high end of the skill and transparency of its financial system, the strength distribution are much more likely to find satisfactory of its infrastructure, its regulatory and rule-of-law employment as well as opportunities to enhance or environment, natural resources, entrepreneurship, broaden their skills over their working lives. In the economic flexibility, monetary and fiscal policies, United States, employer investment in the education and culture all play key roles. In the long run, how- and training of their workforce is strongly linked to ever, a broadly educated and skilled workforce is an

60 Robert E. Rubin, “We Must Change Policy Direction,” The Wall Street Journal, January 24, 2006, p. A-20. 61 The National Center for Public Policy and Higher Education, 2005; Aspen Institute, 1998.

24 the educational attainment and literacy proficiencies While the data from the large-scale national and workers bring with them to the job.62 Because they international surveys show that literacy proficiencies have an adequate foundation, workers who possess appear to have a significant impact on labor-market these proficiencies are able to take full advantage of outcomes over and above years of schooling, they do those opportunities. not address an important macroeconomic question: Do these proficiencies contribute to the overall GDP These different dynamics at the two ends of the and productivity growth of a country? A recent study skills continuum have led to a widening gap between by a group of economists using IALS data from 14 the haves and have-nots. Other factors also contrib- OECD countries found that not only do direct mea- ute to this dynamic. Globalization and the diffusion sures of literacy have a significant and positive effect of information and communication technologies, as on levels of per capita GDP and productivity, they do well as recent trade agreements, have given employers a better job than measures of educational attainment in and corporations more leeway in reducing their com- predicting economic growth. The study found that human mitments to workers, because they are less bound by capital as measured in these surveys is important for national borders.63 This means that a great many indi- societal as well as for individual growth.64 viduals and families are just one crisis away from di- saster. For example, in New Orleans, those with more As we argue in this report, the growth of human resources were better able to weather the onslaught of capital and how that growth is distributed are both Hurricane Katrina. important for America. Human capital is important, however, not just because of the economic implica- It is not likely we will grow our way out of this tions. A recent book by Benjamin Friedman also links problem any time soon. Evidence from NAEP has our economic well-being to the very fabric of our shown little improvement in reading and only mod- society.65 Friedman argues that economic growth and est improvement in math achievement of the nation’s stability are tied to the quality of our democracy. He 13- and 17-year-olds over the past 20 years, and almost shows that the history of the 20th century is replete no change in the gaps between White and minority with examples in which democratic institutions suf- students. What if we are unsuccessful in substan- fered during periods of economic stagnation and tially improving formal educational attainment and downturns. Friedman reasons that individuals and academic achievement over the course of the next 20 societies are more trusting, more inclusive, and more years? When we combine current skill distributions open to change when they see their futures and the of the adult population with current and projected futures of their children as bright and secure. He con- demographic trends to obtain a plausible projection of cludes that “Economic growth is not merely the enabler skill distributions in the future, the result is a nation of higher consumption; it is in many ways the well- characterized by lower average skills and increased spring from which democracy and civil society flow.” variation. Today, skill distributions are mirrored by substantial variations in employment rates, access to America’s perfect storm is greater than the simple skilled jobs, incarceration rates, earnings, incomes, sum of the three sets of forces that are discussed in wealth, health status, degree of civic participation, and this report. The confluence of those forces can create a so on. These gaps prevail across subgroups including powerful dynamic that continually feeds the storm — gender, racial/ethnic, and whether or not individuals putting at risk not only greater numbers of individuals were born in the United States. but the very fabric of our nation. A future reflecting

62 Evidence on this issue is based on the IALS data; see Sum, Kirsch, and Yamamoto, 2004. 63 Jeff Faux, The Global Class War, Hoboken, NJ: John Wiley & Sons, 2006. 64 Serge Coulombe, Jean-Francois Trembly, and Sylvie Marchand, Literacy Scores, Human Capital, and Growth Across Fourteen OECD Coun- tries, Ottawa: Statistics Canada, June 2004. 65 See Benjamin M. Friedman, “Meltdown: A Case Study,” Atlantic Monthly, July/August 2005. For a more comprehensive discussion, see Benjamin M. Friedman, The Moral Consequences of Economic Growth, New York: Knopf, 2005.

25 the projected changes in demographics and skill dis- tributions is one in which there would be fundamental changes in existing economic and social structures. The implicit promise of every individual having a fight- ing chance to improve his or her station in life would be replaced by the reality of what columnist David Brooks has called an “inherited meritocracy.” 66 There are those in business and government who are not troubled by the recent concentration of wealth and power that contrasts with the more broadly shared prosperity America experienced in the decades following World War II. While there is little doubt that America will continue to grow and prosper, the chal- lenge we face, given the forces described in this report, is to decide whether we allow our country’s popula- tions to continue to grow apart or, as a nation, we will invest in efforts to help us grow together. The authors of this report believe the latter course is better for America — and that the time to act is now.

66 David Brooks, “Psst! Human Capital,” The New York Times, November 13, 2005, p. A12.

26 Appendix Tables

Table A1 Table A2 Average Scores of U.S. Adults 16 to 65 Years of Trends in Average NAEP Scores for 13- and Age and Percentages in Levels 3 and Higher on 17-Year-Old Students in Reading and Math the Literacy and Numeracy Scales NAEP Assessment Years NALS ALLS 1984 1990 1996 2004 Mean Percent in Mean Percent in Reading Score Level 3+ Score Level 3+ Age 13 257 257 258 259 Prose 280 58 269 47 Age 17 289* 290 288 285 Document 276 54 270 48 1986 1992 1996 2004 Quantitative 279 57 Math Numeracy 261 42 Age 13 269* 273* 274* 281 Age 17 302* 307 307 307 Source: National Adult Literacy Survey 1992 and Adult Literacy and Life Skills Survey 2005. Tabulations by the authors. Source: Marianne Perie, Rebecca Moran, and Anthony D. Lutkus, NAEP 2004 Trends in Academic Progress: Three Decades of Student Performance in Reading and Mathematics (NCES 2005-464), U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, Washington, D.C., July 2005.

*Significantly different from 2004.

Table A3 Distribution and Average Proficiency of Adults on the ALLS Prose Scale, by Race/Ethnicity and Origin of Birth

Level 1 Level 2 Level 3 Levels 4/5 Mean Prose Score Total Adult Population 20 32 35 13 269 White (total) 12 30 41 17 282 Native 11 30 42 17 284 Non-Native 27 28 34 11 263 Black (total) 33 44 20 3 242 Native 34 43 20 3 242 Non-Native 20 60 17 3 249 Hispanic (total) 49 33 14 3 225 Native 30 38 27 5 252 Non-Native 67 27 5 1 202 Asian (total) 28 33 30 9 256 Native 5 21 52 22 297 Non-Native 29 34 30 7 252

Source: Adult Literacy and Life Skills (ALLS) Survey 2005. Tabulations by the authors.

27 Table A4 Table A5 Trends in Average White-Black and White-Hispanic Changes in the Number of Employed Persons Differences Among 13- and 17-Year-Old Students (Age 16+), Total and by Major Occupational Group, in Reading and Mathematics 1984-2000: United States (Annual Averages, Numbers in 1,000s) 1984 1994 2004 Reading Occupational Absolute Percent Group 1984 2000(1) Change Change White-Black All 16+ 105,005 135,208 30,203 28.8 Age 13 26 31* 22 Professional 13,286 21,113 7,827 58.9 Age 17 32 30 29 Managers, execu- 11,571 19,774 8,203 70.9 White-Hispanic tives, and manage- Age 13 23 30* 24 ment support Age 17 27 33 29 Technical 3,172 4,385 1,213 38.2 1986 1996 2004 Sales Math • High-level 6,685 9,452 2,767 41.4 White-Black • Lower-level(2) 5,897 6,889 992 16.8 Age 13 24 29 27 Administrative 16,722 18,717 1,995 11.5 support Age 17 29 27 28 Service workers 14,151 18,278 4,127 29.2 White-Hispanic Farm, forestry, 3,600 3,399 -201 -5.6 Age 13 19 25 23 fishing Age 17 24 21 24 Precision produc- 13,057 14,882 1,825 14.0 tion, craft/installa- Source: Marianne Perie, Rebecca Moran, and Anthony D. Lutkus, NAEP 2004 Trends in tion, repair Academic Progress: Three Decades of Student Performance in Reading and Mathematics (NCES 2005-464), U.S. Department of Education, Institute of Education Sciences, National Production work- 12,451 12,876 425 3.4 Center for Education Statistics, Washington, D.C., July 2005. ers and transport * Significantly different from 2004. operatives Laborers, helpers, 4,413 5,443 1,030 23.2 cleaners

Source: U.S. Bureau of Labor Statistics, Employment and Earnings January 1985 and Employment and Earnings January 2001.

(1) The occupational employment estimates for calendar year 2000 are based on the popu- lation weights from the 1990 Census. These employment estimates were later revised by Bureau of Labor Statistics to reflect new population weights from the 2000 Census, but they did not revise the occupational employment estimates under this coding system.

(2) Lower-level sales workers include sales clerks, cashiers, counter clerks, door-to-door salespersons, and newspaper vendors.

28 Table A6 Table A7 Estimated Changes in Civilian Employment in Se- Projected Changes in Occupational Employment lected Major Occupational Groups, United States: in the United States, in Selected Occupational 2000-2005 (Numbers in 1,000s, Annual Averages) Groups, 2004-2014 (in 1,000s, Annual Averages)

Occupational Absolute Percent Occupational Absolute Percent Group 2000 2005 Change Change Group 2004 2014 Change Change All 137,101 142,127 5,026 3.7 All 145,612 164,540 18,928 13.0 Management 14,348 14,707 359 2.5 Management, 14,987 17,142 2,155 14.4 occupations business, and Business and fi- 5,277 5,771 493 9.4 financial nancial operations occupations Professional and 26,740 28,835 2,096 7.8 Professional and 28,544 34,590 6,064 21.2 technical, all technical, all • Architecture and 2,990 2,794 -196 -6.6 High-level sales 5,325 5,818 466 8.7 engineering Retail Sales 8,445 9,382 937 11.1 • Computer and 3,318 3,245 -73 -2.2 Service 27,673 32,930 5,257 19.0 mathematical occupations science Office and admin- 23,907 25,287 1,380 5.8 • Education, train- 7,239 8,129 891 12.3 istrative support ing, and library Construction 7,738 8,669 931 12.0 • Health care 5,909 6,759 850 14.4 and extraction practitioner and Installation, 5,747 6,404 657 11.5 technical maintenance, Management, 46,365 49,313 2,948 6.4 and repair professional, Production 10,562 10,483 - 79 - .7 and technical combined Transportation and 10,098 11,214 1,116 11.1 material moving Source: 2000 and 2005 Current Population Surveys, public use files. Tabulations by the authors. Source: Daniel Hecker, “Occupational Employment Projections to 2014,” Monthly Labor Review, November 2005.

Table A8 The Contributions of Net International Migration to U.S. Population Growth, Selected Time Periods, 2000 to 2005

Net Immigration as a Net International Percent of Population Time Period Population Growth Migration Change April 2000-July 2005 (actual) 14,985,502 6,333,941 42 2005-2010, projected (1) 13,428,447 6,727,438 50 2010-2015, projected (2) 13,430,206 7,124,863 53

Source: U.S. Census Bureau, www.census.gov.

Note: (1) Sum of net international migration is for the 2006-2010 period only, since new immigrants in 2005 are included in the population count for that year. (2) Sum of net international migration is for the 2011-2015 period only, since new immigrants in 2010 are included in the population count for that year.

29 Table A9 Table A11 Listing of Countries Sending 100,000 or Percentage of 36- to 43-Year-Olds Who Have 16 or More Immigrants to the United States More Years of Schooling, by Gender, Race/Ethnicity, Between 2000 and 2004 (All Ages) and Decile of the AFQT Distribution (2002)

Country Number of Immigrants Decile All Men Women White Black Hispanic Mexico 2,170,794 Total 26.5 25.8 27.2 30.3 14.9 14.6 India 345,063 First 1.6 1.7 1.4 0.0 2.6 0.8 China 235,115 decile Philippines 202,314 Second 2.2 1.7 2.7 0.7 4.3 2.6 decile Puerto Rico* 185,364 Third 4.3 4.0 4.7 2.7 8.3 5.3 El Salvador 160,712 decile Guatemala 129,131 Fourth 8.8 6.4 11.0 6.3 15.1 12.9 Brazil 120,223 decile Colombia 112,795 Fifth 8.8 9.8 8.1 7.0 18.4 10.0 Canada 109,962 decile Cuba 109,939 Sixth 15.4 9.0 20.9 13.6 26.5 19.9 decile Japan 109,898 Seventh 21.7 18.2 24.8 20.8 36.2 22.2 Dominican Republic 107,283 decile Source: 2004 American Community Surveys, public use files. Tabulations by the authors. Eighth 32.8 31.5 33.9 33.1 44.1 33.9

* The authors recognize that Puerto Rico is a U.S. territory with com- decile monwealth status, but the island does have a language and culture of its own. Data show that the island’s residents who come to the mainland Ninth 51.4 45.1 58.1 50.4 73.6 62.5 experience some of the same challenges as immigrants who come to the decile United States from other countries. In addition, the United States does not include residents of Puerto Rico in its official population counts of Tenth 76.2 75.8 76.7 76.5 72.1 79.5 labor force numbers. A resident of Puerto Rico who migrates to the United States adds to the nation’s population similar to that of an decile immigrant from another country. Source: National Longitudinal Survey of Youth. Tabulations by the authors.

Table A10 Percentage of Adults Ages 16 to 65 in Each Skill Level, 1992 NALS and Projections for 2030

1992 NALS Data Projected Distribution 2030 Prose Document Quant Prose Document Quant Level 1 17 19 18 27 26 26 Level 2 26 28 26 27 27 25 Level 3 35 34 33 28 28 28 Level 4 19 17 19 14 14 16 Level 5 3 2 4 4 4 5 Mean (sd) 280 (63) 275 (63) 279 (67) 266 (68) 262 (68) 266 (71)

Source: National Adult Literacy Survey data and authors’ projections.

30 Table A12 Mean Lifetime Earnings of 18- to 64-Year-Old Males in the United States, by Educational Attainment, (in Constant 2005 CPI-U Dollars)

Educational Attainment (A) (B) (C) (D) 1979 2004 Absolute Change Percent Change No high school diploma or GED $1,577,466 $960,365 -$617,101 -39.1 H.S. diploma/GED, no completed years of college $1,814,595 $1,380,636 -$433,957 -23.9 1-3 years of college, including Associate’s degree $2,007,712 $1,738,411 -$269,301 -13.4 Bachelor’s degree $2,736,270 $2,702,793 -$33,478 -1.2 Master’s or higher degree $3,039,355 $3,506,939 $467,584 15.2 All $1,879,696 $1,902,375 $22,679 1.2

Source: 1980 Census of Population and Housing, PUMS files, tabulations by the authors; and2004 American Community Surveys, public use files, tabulations by the authors.

Table A13 Relative Mean Annual Earnings of U.S. Adults* in Selected Educational Attainment and AFQT Decile Groups, 1988, 1997, 2001

Year HS Grad to Some College B.A.+ to HS B.A.+ Avg Skills to B.A.+ Top Skills to B.A.+ Low Skills to HS Dropout to HS Grad Grad HS Grad Avg Skills HS Grad. Top Skills HS Grad Low Skills 1988 1.38 1.20 1.58 1.01 1.69 1.14 1997 1.31 1.14 1.81 1.35 1.68 1.18 2001 1.35 1.17 2.01 1.47 2.22 1.29

Sources: National Longitudinal Survey of Youth (NLSY 1979), 1989, 1998, 2002 survey records. Tabulations by one of the authors.

Note: Average skills defined as AFQT in 5th to 6th decile; top skills as AFQT in 9th to 10th decile; and low skills as AFQT in 2nd to 3rd decile.

* The sample of adults were 23- to 30-years-old in 1988 and 36- to 43-years-old in 2001.

Table A14 Mean Weekly Earnings of U.S. Full-Time Employed Adults (Age 16 and Older) by Prose Proficiency Level and Educational Attainment

Educational Attainment Level 1 Level 2 Level 3 Level 4 Level 5 Overall 0 to 8 years $298 $351 ------$313 9 to 12 years $364 $357 $414 ------$373 GED $333 $364 $489 $529 --- $431 High school diploma $369 $420 $436 $493 --- $430 Some postsecondary $367 $455 $491 $597 --- $509 Two year degree $386 $504 $578 $610 $630 $574 Four year degree or higher $586 $677 $739 $866 $993 $830

Source: NALS Survey, 1992.

Note: --- Indicates that the number of cases is too small to provide reliable estimates.

31 Table A15 Table A17 Total Births to Women Under Age 30, and Percentage of Births to Women Under Age 30 Number and Percentage That Were Out of That Were Out of Wedlock, by Educational Wedlock, by Race/Ethnicity of Mother, Attainment and Race/Ethnicity of the Mother, United States: 2003-2004 United States: 2003-2004 Out-of- Percent Educational Black White Hispanic Total Wedlock Out of Attainment Mothers Mothers Mothers Group Births Births* Wedlock <12 or 12, no 89.9 60.7 50.0 All 2,298,318 1,011,187 44.0 diploma Asian 77,173 12,586 16.3 12 years, 80.0 42.1 48.6 Black 381,738 291,846 76.5 diploma or GED Hispanic 525,799 242,622 46.1 13 to 15 years 69.1 28.0 38.6 Mixed Race 335,517 19,972 60.0 Bachelor’s degree 41.0 8.0 28.0 White, not Hispanic 1,242,260 420,467 33.8 Master’s degree 11.3 3.3 15.3 or higher Source: 2004 American Community Surveys, public use files. Tabulations by the authors. *Out-of-wedlock births are modestly underreported on the ACS surveys. Source: 2004 American Community Surveys. Tabulations by the authors.

Table A16 Number and Percentage of Births to Women Under Age 30 That Were Out of Wedlock, by Educational Attainment of the Mother, United States: 2003-2004 Out-of- Percent Educational Total Wedlock Out of Attainment Births Births* Wedlock <12 or 12, no 563,288 350,847 62.2 diploma 12 years, 743,151 381,060 51.3 diploma or GED 13 to 15 Years 666,328 242,920 36.7 Bachelor’s degree 268,238 33,669 12.6 Master’s degree 62,313 2,691 4.3 or higher All 2,298,318 1,011,077 44.0

Source: 2004 American Community Surveys, public use files. Tabulations by the authors *Out-of-wedlock births are modestly underreported on the ACS surveys.

32 About ETS

ETS is a nonprofit institution with the mission to advance quality

and equity in education by providing fair and valid assessments,

research and related services for all people worldwide. In serving

individuals, educational institutions and government agencies

around the world, ETS customizes solutions to meet the need

for teacher professional development products and services,

classroom and end-of-course assessments, and research-based

teaching and learning tools. Founded in 1947, ETS today develops,

administers and scores more than 30 million tests annually in

more than 180 countries, at over 9,000 locations worldwide.

Additional information about the America’s Perfect Storm report

is available at ets.org/stormreport.

741355

7707